Will ChatGPT undermine ethical values in nursing education, research, and practice?

Abstract

ChatGPT is an artificial intelligence-based chatbot that uses deep learning techniques to generate natural language text (ChatGPT, 2022). While ChatGPT can enhance efficiency, there are growing concerns about its safety and future implications for humanity. On March 29, 2023, industry leaders in artificial intelligence (AI) signed a petition to pause research into AI that is more powerful than the ChatGPT4. According to the petitioners, this action was in response to the uncontrolled “race to develop and deploy ever more powerful digital minds that no one—not even their creators—can understand, predict, or reliably control” (Future of Life Institute, 2023, para. 1). Indeed, AI leaders are concerned about the “significant risk to humanity” that could result from an uncontrolled AI tools that lacks the necessary safety regulations. As nursing researchers in digital health technology and AI, we are concerned not only about the safety implications of ChatGPT, but also about its ability to capture the ethical values, principles, and core tenets that underpin the unique discipline of nursing. Recent editorials and commentaries in some nursing journals have discussed the potential benefits, limitations, and risks of using ChatGPT in nursing education and practice (Archibald & Clark, 2023; Odom-Forren, 2023; Scerri & Morin, 2023). In this commentary, we extend upon these discussions by reflecting on how the use of ChatGPT may undermine the values, principles, and core assumptions that underpin nursing research, education, and practice. We aim to stimulate a dialogue around current and emerging trends in using ChatGPT in nursing. We also want to engage nursing scholars, educators, and practitioners across international communities to discuss nursing assumptions in the context of technology use. Open AI acknowledged that ChatGPT cannot ensure the confidentiality of information entrusted onto it. Based on this, open AI advises against the inclusion of potentially sensitive, confidential, and identifiable information. The limitations of ChatGPT in processing confidential information may undermine the privacy and confidentiality inherent in nursing practice. Nurses are entrusted with clients' personal and health information and they have a legal and ethical responsibility to protect such information at all times. Compromising such information using Open AI tools like ChatGPT might endanger the nurse–client relationship, thus creating distrust in healthcare systems (Scerri & Morin, 2023). Aside from the potential for creating distrust in nurse–patient relationships, we are also concerned that ChatGPT might not capture human emotions, such as empathy and compassion, that are central to the core discipline of nursing. It is important to note that nursing is value-based, emphasizing compassion, empathy, care, respect, and dignity for the patient. On the other hand, Open AI prioritizes the practical usefulness of computing over offering explanatory or theoretical analysis. Without attuning to nursing disciplinary perspectives, ChatGPT may not be able to integrate the unique perspectives of nursing into the ChatGPT algorithms. With ChatGPT becoming common in the healthcare field (Sallam, 2023), it is important to critically reflect upon the potential and problematics of mixing AI data with a humanistic discipline like nursing, especially when they do not fit well. Another point of reflection is the use of ChatGPT in light of the decision-making process in Nursing. ChatGPT uses deep learning algorithms to process sequences of data, learn patterns in the data and generate text based on a query (ChatGPT, 2022). In effect, the output of ChatGPT is based on data available at any point in time. On the other hand, decision-making and clinical judgment in nursing go beyond data and include intuition, tacit knowledge, wisdom, and personal experiences. Even though ChatGPT may be designed to learn, reason, and solve problems that mimic human cognition, it may not be possible to identify or solve nursing problems that demand wisdom, critical thinking, intuition, and personal judgment. To illustrate, the American Nurses Association has included wisdom in the Data-Information-Knowledge framework of computing systems—changing the framework to Data-Information-Knowledge-Wisdom (Nelson, 2020). While AI tools like ChatGPT may be able to represent and generate information relevant to the first three dimensions, wisdom has been a more elusive concept that has not been adequately represented in computerized systems (Matney et al., 2011). This suggests that an overreliance on ChatGPT means a gradual disengagement in critical thinking, wisdom, intuition, tacit knowledge, and ethical judgment, which are considered attributes that define professional nursing practice, education, and research. The complexity and particularity inherent in nursing also constitute ethical values that may not sit well with the use of AI tools like ChatGPT. A recent blog post by a nursing student indicated how ChatGPT can aid nurses' understanding of complex concepts by breaking down complex ideas into simpler pieces (Tran, 2023). It is important to note that nursing adopts a holistic approach to care and clients cannot simply be understood by breaking the whole into parts as this blog post seems to suggest. Nursing considers each individual client as unique and complex. The reductionist approaches of ChatGPT may be overly simplistic and may not be capable of analyzing or understanding clients that present with complex health challenges or tailoring responses to individual cases. Therefore, ChatGPT-generated responses may not fully capture the holistic knowledge needed to address the complex contexts and nuances of individual health challenges. The reductionist approaches of ChatGPT may also conflate the disciplinary boundaries between nursing and medicine. Current AI applications specific to medicine are increasingly applied in disease diagnosis, personalized disease treatment, medical imaging, and genomic sequencing (Amisha et al., 2019). AI may be common in Medicine because AI applications are programmed to function in ways that are similar to the biomedical model of the human body (Kumar et al., 2022). Rather than focusing on individual ailments, nursing disciplinary boundaries transcend the biomedical model of care to focus on the broader picture of the patient, including the psychological, emotional, and social factors that come together to determine one's health. Therefore, Adopting AI tools like ChatGPT in nursing education and practice may be akin to adopting biomedical approaches in situations that demand psycho-sociological approaches to care. Another major concern of ChatGPT in nursing research and education relates to using the tool as an author and how that practice can facilitate academic dishonesty by enabling students to generate written assignments, discussion posts, or research papers without engaging in original thought or research (Choi et al., 2023). ChatGPT was recently listed as a coauthor in published peer-reviewed nursing papers (ChatGPT Generative Pre-Trained Transformer & Zhavoronkov, 2022; O'Connor & ChatGPT, 2023). Such practice raised serious questions about human and AI authors as the latter do not have accountability as a human author (Stokel-Walker, 2023). This practice of generating content using ChatGPT undermines the moral value of integrity in nursing, which demands nurses to be honest with themselves, their patients, and other healthcare teams. Furthermore, nursing decisions involve moral and value judgments and reliance on ChatGPT in nursing education may produce nurses with less adherence to moral values, little or no sense of trustworthiness, and overconfidence in handling complex cases. While industry experts have called for a pause to AI research, the reality is that the development of AI models will continue to advance and these may well find their way into nursing practice, education, and research. Until robust regulatory mechanisms on AI are developed, nursing must adopt some actionable strategies on ChatGPT that will ensure that the use of the AI tool does not distort the ethical values of nursing practice, education, and research. For instance, nursing educators may consider working with students/learners/trainees to establish clear guidelines and ethical standards that govern the use of AI tools like ChatGPT in nursing education, research, and practice. Additionally, promoting a culture of academic integrity and raising awareness of the ethical implications of AI use can help mitigate against academic dishonesty while preserving integrity and honest nursing practice. To counteract disengaged critical thinking and ensure holistic nursing practice that responds to complex situations, educators should prioritize the development of critical thinking and problem-solving skills in nursing curricula. This can be achieved by incorporating activities that encourage students to critically analyze and evaluate information, synthesize knowledge from various sources, and apply their understanding. Nursing academics should also emphasize the importance of contextualized and individualized decision-making in nursing by incorporating case studies, simulations, and real-life scenarios that highlight the importance of patients' unique needs and complex circumstances. Developing clear guidelines and promoting academic integrity, fostering critical thinking skills and independent problem solving, and encouraging contextualized and individualized decision-making can help ensure that AI tools like ChatGPT are used responsibly and ethically, ultimately enhancing the quality of nursing care practice, research, and education. Using ChatGPT in nursing poses ethical challenges that could undermine nursing's core values at the same time that integrating AI in nursing research, education, and practice offers potential advancement. As AI tools continue to grow, it is vital for nursing scholars and educators to engage in critically reflective dialogues and philosophical analyses of the implications, ethics, and potential drawbacks associated with using AI in nursing contexts. By critically questioning the underlying assumptions of AI-driven innovations, we can ensure that the evolving role of AI in nursing aligns with the values, principles, and goals traditionally shaping the profession, promoting holistic, person-centered, and ethical healthcare outcomes. Failing to do so, we risk eroding the foundations that make nursing a unique and essential discipline. No new data were generated for this commentary.

Ähnliche Arbeiten